Reactive throttle infrastructure

A method, comprising:

• receiving a request from a client to run a task utilizing a network resource, wherein the client accesses the resource to run the task through multiple ones of a plurality of logical lanes, wherein the plurality of logical lanes represent channels through which the client interacts with the resource, and wherein each of the plurality of logical lanes is configured to accept requests from a different class of clients including the client such that different service requests are allocated to different lanes based on the service type of the service request in accordance with the class of client associated with the service;
• analyzing the request and criteria associated with the request based on a permit issuance policy to determine a condition for granting the permit, wherein the condition includes one or more queue limits associated with one or more of the plurality of logical lanes, wherein each of the lanes is capable of having a different independent queue limit from each of the other of the plurality of logical lanes, and wherein the permit is for one of the plurality of logical lanes;
• granting the permit for the one of the plurality of logical lanes to the client based on satisfaction of the condition;
• effecting running of the task based on granting of the permit and interacting with the resource through the one of the logical lanes;
• monitor resource performance associated with performance of the service requests;
• and
• defining a throttling policy based on a comparison of the monitored resource performance to one or more performance thresholds of the resource and traffic for each of the plurality of logical lanes associated with the resource.